More than 40 chromosomal regions harboring risk loci for T1D have been mapped through genome-wide association studies (GWAS). However, for most of these regions, the causative genetic variants, and the genes they act upon, have yet to be identified. As germline factors, genetic risk variants are present and amenable to study at all times - before, during and after the development of diabetes. Therefore, genetic information can serve as a potential predictive tool as well as provide insights into pathogenesis occurring during the preclinical phase of the disease where preventive therapies might be applied. In this proposal we describe an approach for identifying T1D risk variants and characterizing their function that focuses on T1D families with 3 or more affected siblings characterized by early ages at onset. In these unusual high risk T1D families, we use sequencing to identify rare, overtly deleterious variants affecting genes located in T1D risk regions defined in prior GWAS studies. Our premise is that any gene that contributes to disease risk is unlikely to have just one common risk variant, but rather should contain multiple risk alleles with a range of frequencies. Identifying risk alleles that, while less frequent, have more overt and readily discernible effects on gene function can both pinpoint the relevant gene from among many in a T1D-associated chromosomal region and provide insights into its mode of action in T1D pathogenesis. In preliminary studies we have used this approach to identify a novel T1D-associated deleterious variant in UBASH3A which affects the splicing of the gene, resulting in the production of a unique isoform of the protein that differs in function from the product of the wild type allele. Ou preliminary data reveal a significant role for UBASH3A in human T cell activation not predicted from prior mouse knockout studies. The proposed studies build upon these findings in two ways. First we pursue mechanistic studies of UBASH3A in order to define its function in human T cells, understand its previously unrecognized role in the nucleus and define its contribution to T1D risk. To this end, we have generated CRISPR based knockouts of UBASH3A in Jurkat cells allowing the introduction of specifically mutagenized constructs to test the effects of mutations on cell activation. In parallel, we propose to expand our genomic approach that led to these findings, extending whole exome sequencing to additional high risk T1D families, improving our filtering approaches for causative variants and identifying new genes contributing to T1D pathogenesis which can be targeted for functional studies.